← Changelog

Image-to-video, DaVinci Resolve, and the full creation stack

Image-to-videoDaVinci ResolveTimelineAPIChatGPT
  • Image-to-video with sync-3 - Turn a still image into a naturally speaking video. Upload a JPEG, PNG, or WebP image, add audio, choose the face you want to animate, and let sync-3 build the performance. Available in the App, API, Adobe Premiere Pro, and DaVinci Resolve.
  • DaVinci Resolve plugin - Bring sync. labs into DaVinci Resolve. Use footage from your timeline or media pool, generate a result, and send it straight back to your project. Download the plugin
  • More control on the Timeline - Crop individual clips, undo and redo edits, trim with frame-level precision, and keep full-resolution output through the edit. The preview now matches the final render more closely, so you can make decisions before spending a generation.
  • The full creation stack, now in the API - Upload and reuse assets, synthesize speech, clone and manage voices, organize work into projects, generate lipsync, and wait for results in one request. Explore the API
  • Create directly in ChatGPT - Upload a video or image, create speech, clone a voice, animate an image, generate lipsync, organize projects, and retrieve results without leaving the chat. Use sync. labs in ChatGPT
Improvements
  • Download every completed generation in a project as one ZIP, including work created across the App and API.
  • Preview an asset by hovering over it and pressing Space.
  • The App now warns before generation when multiple speakers are detected and no speaker has been selected.
  • Closed-lip footage now prompts you to use sync-3, which handles that starting position more naturally.
  • Compare original and dubbed audio inside the dubbing preview before generating the final video.
  • The App reads professional media metadata more reliably for MOV, ProRes, MXF, and DNxHD files.
  • Adobe Premiere Pro and DaVinci Resolve can submit another generation while earlier uploads continue in the background.
  • Generation errors now distinguish invalid requests, duplicate submissions, timeouts, unavailable services, and concurrency limits, with clearer guidance on what to do next.
Bug fixes
  • Fixed sync-3 rejecting some videos that were within the duration limit after sync mode was applied.
  • Fixed generated text-to-speech audio not appearing in the preview.
  • Fixed the App occasionally staying on the Source view after a generation finished.
  • Fixed the Timeline going empty while a generation was processing.
  • Fixed audio clips being placed beyond the end of the video track.
  • Fixed intermittent Adobe Premiere Pro plugin failures during project scanning, uploads, and authentication.